HAL Id: hal-02800279
https://hal.inrae.fr/hal-02800279
Submitted on 5 Jun 2020
HAL is a multi-disciplinary open access archive for the deposit and dissemination of sci- entific research documents, whether they are pub- lished or not. The documents may come from teaching and research institutions in France or abroad, or from public or private research centers.
L’archive ouverte pluridisciplinaire HAL, est destinée au dépôt et à la diffusion de documents scientifiques de niveau recherche, publiés ou non, émanant des établissements d’enseignement et de recherche français ou étrangers, des laboratoires publics ou privés.
La validation de méthode d’analyses comme outil d’évaluation de la fiabilité des résultats de recherche
Elodie Ollivier, Anne Jaulin, Virginie Grondin, Amelie Trouve, Nathalie Cheviron
To cite this version:
Elodie Ollivier, Anne Jaulin, Virginie Grondin, Amelie Trouve, Nathalie Cheviron. La validation de méthode d’analyses comme outil d’évaluation de la fiabilité des résultats de recherche. Les 14èmes Journées de la Mesure et de la Métrologie, Oct 2016, La Chaussée-St-Victor, France. 2016. �hal- 02800279�
THE VALIDATION OF ANALYTICAL METHODS AS AN EVALUATION TOOL OF RESEARCH RESULTS RELIABILITY
Ollivier Elodie, Jaulin Anne, Grondin Virginie, Trouvé Amélie, Cheviron Nathalie
Context
UMR ECOSYS INRA, AgroParisTech, Université Paris-Saclay, plateforme Biochem-Env, 78026, Versailles, FRANCE
Conclusions
Validation of analytical methods by the accuracy profile approach
Results
The platform Biochem-Env:
Was created in 2012 by INRA (French National Institute for Agricultural Research) with the support of the ANR program “Investissements d’avenir” as a service of the infrastructure ANAEE-France, For the biochemical characterization of natural environments (soils and sediments) and associated macrofauna in research projects,
By developing and validating methods in order to provide traceable analytical data with high level of confidence.
For intra-laboratory validation of quantitative analytical methods, the INRA’s Quality Guidelines for research and experimental units (2013) recommends “the accuracy profile” method according to the NF V03-110:2010 standard.
Could we use a same internally developed method to quantify proteins in various biological models ?
Material and methods
The validation of this analytical method helped us to:
- Determine the performance of the method
- Improve the steps for sample preparation and analysis - Assess the matrix effect
- Pointed out the importance of an experimented analyst
Analytical method for protein quantification in 4 biological models was validated with a good accuracy considering scientific specification and needs.
Accuracy profile approach:
- Global statistical combination of trueness and precision and pragmatic
- Simple graphic interpretation, allowing a clear and easy comparison between method performance and intended use and, a rapid decision
- No limit in the choice of the calibration model => large scope range
- Methods with very low variability can be validated (not rejected by a H0) - Diagnostic tool, matrix effect taken into account
- Risks and guarantees managed for both end-users and laboratories - Estimation of measurement uncertainty
Purpose:
To provide guarantees on analytical results, for the analyst and the end-user
To demonstrate analytical method fitting with the scientific objectives
To allow laboratory recognition
To improve analysts working practices
Benefits of the accuracy profile approach:
An overall statistical method combining trueness and precision A standardized approach : NF V03-110:2010
A simple and graphic interpretation for a rapid decision The determination of the scope of the method
The determination of quantification limits An estimation of measurement uncertainty
Protein determination method by the Bicinchoninic acid (Ref. QPBCA-1KT, QuantiPro™ BCA Assay Kit, Sigma-Aldrich)
Aporrectodea icterica (earthworm) Poecilus cupreus (beetle)
Gammarus pulex (gammarid) Alopecosa pulverulenta (spider)
Validation method according to the « accuracy profile » approach (NF V03- 110:2010 standard):
Selection of validation samples:
- Spiked reference material: Bovine Serum Albumine - 7 levels (0.5 ; 1 ; 2 ; 5 ; 10 ; 20 ; 30 mg.mL-1)
- 3 replicates (repeatability)
- 6 days of analysis (intermediate precision) - 2 analysts (intermediate precision)
Calibration (indirect quantitative analysis):
- 6 levels (0 ; 2 ; 4 ; 8 ; 16 ; 20 ; 30 ; 40 mg.mL-1) - Regression model : polynomial
x x x x xm x x x x x
xv
Trueness (method) Precision
(method)
Accuracy (result)
x : measures
xm : mean value of measurements xv : true value
Definition of the needs
Optimisation of the method Scientific approach
Choice of the method
Research question
Program analysis Performances
Interpretation
Characterisation of the method
Routine analysis Quality control
Results
Validation of analytical method
Estimation of measurement uncertainty
Question : Is this quantitative analytical method fit for purpose in terms of trueness and precision ?
λ-acceptability limits fixed at ±10%
Conclusion :
Data are obtained with internal reference material (spiking):
• True and precise method between 0,4 et 2,4 mg/L (scope of the method)
• Low limit of quantification (LQ) = 0,4 mg/L
• The method fits to the scientific needs for a direct application with a measurement uncertainty estimated at maximum 5% in the field of the validated method.
In the most critical conditions of the scope of the method (low LQ), for each future sample, the method will provide results with the required accuracy. In the wider field of application, the likelihood of obtaining non
acceptable results is very low.
β-expectation tolerance limits fixed at ±80%
4 biological models
Biological model effect Matrix effect
Analyst effect
Performance of the method
Limit of Quantification LQ:
2,7 mg/mL
Scope of the method:
2,7 to 30 mg/mL
Measurement uncertainty range (U95%) : 10 to 30%
Fixed criteria:
β= ±80% (expectation tolerance limits)
λ= ±20% (acceptability limits)
1 analyst 2 analysts
Decision on the fitness for
purpose
- Adapt λ-acceptability values according to the concentration range
- Extend the method to other biological models and biomarkers (Lipid, glycogen…)
Biochem-Env
Perspectives :
Correction factor application